The Second AI Winter and Resurgence of AI During
1980-2010

Although Deep Blue garnered enormous headlines in 1997, progress in AI during 1980-2010 went through peaks and valleys. During the early 1980s, Expert Systems, which had emerged as one of the sub-fields of AI in the 1960s, attracted substantial attention with AI researchers believing it could be the panacea for all their problems. However, those dreams didn’t materialize thereby leading to a “small AI winter” between 1987 and 1993. During the mid-1990s, research funding ebbed and flowed, and then slowly research in AI began to gather steam although “some computer scientists and software engineers would avoid the term artificial intelligence for fear of being viewed as wild-eyed dreamers”.

Despite a few setbacks mentioned above, during 1980-2010, researchers continued to expand the paradigm related to Machine Learning algorithms that were established during 1950-1979. Several of them realized that many Machine Learning algorithms could be improved by using techniques from mathematics and statistics, economics, game theory, stochastic modeling, classical numerical methods, and operations research. Others developed better mathematical descriptions for Deep Learning Networks (DLNs) as well as for evolutionary and genetic algorithms. These techniques gradually began to mature by 2011 and their hard work began to pay off, especially because of several external factors, including computers becoming much faster and cheaper. Eventually, all these efforts set the stage for explosive growth in AI systems starting in 2011. This chapter is partitioned as follows: